Segmenting brain tumours is a critical part of medical image processing. Early detection of brain tumours is critical for enhancing treatment options and boosting the overall survival percentage of patients. When it comes to cancer diagnosis, manual segmentation of the brain tumours from huge amounts of MRI data obtained in clinical practise may be complex and time intensive. Automatic tumour picture segmentation is required. This article provides a basic introduction to the brain and the many disorders that might arise in it. This paper clearly outlines the differences between benign and malignant brain abnormalities. Conventional methods for detecting and sizing brain tumours are also discussed in this study Techniques for diagnosing and treating a brain stroke are also discussed in detail. For the identification of brain tumours and strokes, present algorithms have certain limitations.